An efficient transformation of (-)-quinic acid into carba-L-rhamnose

Article abstract of DOI:10.1055/s-2004-834825

A novel stereoselective synthesis of carba-L-rhamnose from (-)-quinic acid is described. The title compound is obtained with an appropriate pattern of protecting groups to be used for the preparation of Streptococcus pneumoniae type 19F capsular polysacch

Full text of DOI:10.1055/s-2004-834825

LETTER
2529
An Efficient Transformation of (–)-Quinic Acid into Carba-L-rhamnose
T
ransformation
l
of (–)-
e
A
ci
n
C
arba
-
a
L
-rhamnose Valsecchi,a Alessio Tacchi,a Davide Prosperi,b Federica Compostella,c Luigi Panza*a
a
DISCAFF, Università del Piemonte Orientale, via Bovio, 6, 28100 Novara, Italy
Fax +39(0321)375821; E-mail: panza@pharm.unipmn.it
b
c
Istituto di Scienze e Tecnologie Molecolari – CNR, via Golgi 19, 20133 Milano, Italy
Dipartimento di Chimica, Biochimica e Biotecnologie per la Medicina, Università di Milano, via Saldini, 50, 20133 Milano, Italy
Received 29 July 2004
The polysaccharide, however, due to the presence of the
phosphate on the anomeric position is somewhat labile,
making the preservation of a vaccine during the storage
more problematic. Moreover, the presence of the phos-
phodiester bridge makes the chemical synthesis of oligo-
mer of the repeating unit very difficult, not only because
of the lability of the anomeric phosphate, but also for the
possibility to form two anomers for each phosphodiester
bridge.
Abstract: A novel stereoselective synthesis of carba-L-rhamnose
from (–)-quinic acid is described. The title compound is obtained
with an appropriate pattern of protecting groups to be used for the
preparation of Streptococcus pneumoniae type 19F capsular
polysaccharide repeating unit.
Key words: carba-L-rhamnose, stereoselective synthesis, capsular
polysaccharide, quinic acid, shikimic acid
Pneumococcal invasive infections are still a major
problem1 both in developed and developing countries2
causing meningitis, otitis, septicemia and pneumonia. It
would be therefore desirable to have access to vaccines to
protect population against these pathogens.
For these reasons, we decided to synthesize analogs of the
natural repeating units, which are able to avoid the above-
mentioned problems. It has to be highlighted that it is
quite uncommon to use antigen mimics for vaccine prep-
arations. In the case of analogs, which are able to mimic
precisely enough the natural antigen it is reasonable to ex-
pect that the induced response would be able to protect the
organism against bacterial infections.
Among the different pathogens responsible for such infec-
tions, Streptococcus pneumoniae3 is one of the most viru-
lent. S. pneumoniae belongs to the class of encapsulated
bacteria, and antibodies directed against the capsular
polysaccharides (CPS) mediate the resistance to infec-
tions from these pathogens, making these polymers useful
for vaccine preparations, especially using conjugate
vaccines able to induce T-cell dependent response.4
In order to obtain a stable analog of the anomeric phos-
phate, we considered the synthesis of a carba sugar ana-
log, namely carba-L-rhamnose, as it should give a stable
phosphate at the former anomeric position, without prob-
lems with diastereomeric mixture formation. Moreover, it
is possible to consider the possibility to use automated
techniques similar to those for the synthesis of oligo-
nucleotides in order to obtain oligomers with different
chain length (Figure 1).
We are currently involved in a project devoted to the de-
velopment of glycoconjugated vaccines through synthetic
methods. We focused our attention on S. pneumoniae 19F
whose CPS is composed of trisaccharide repeating units
[→4-b-D-ManpNAc-(1→4)-a-D-Glcp-(1→2)-a-LRhap-
1→; Figure 1, a] joined by phosphodiester bridges.
To the best of our knowledge, only one example of the
synthesis of carba-L-rhamnose appeared in the literature.5
The paper describes the formation of carbasugars ex-
O
P
O
P
O
O
O-
O-
CH2
HO
O
HO
HO
HO
O
O
AcNH
O
HO
O
AcNH
O
HO
HO
O
HO
HO
O
HO
HO
HO
O
O
O
OH
OH
b
a
Figure 1 Structure of S. pneumoniae 19F repeating unit (a) and its carba-L-rhamnose containing analog (b).
SYNLETT 2004, No. 14, pp 2529–2532
2
2
.1
1
.2
0
0
4
Advanced online publication: 20.10.2004
DOI: 10.1055/s-2004-834825; Art ID: G29504ST
© Georg Thieme Verlag Stuttgart · New York

2530
E. Valsecchi et al.
LETTER
ploiting a radical cyclization of iodohept-1-enitols de- We decided to use benzyl ethers as permanent protecting
rived from aldoses. Although the approach is attractive, groups for the hydroxyls in positions 3 and 4, to introduce
the aim of such work was the study of the radical cycliza- a naphthylmethyl group,6 orthogonal to benzyls, in posi-
tion, and not the preparation of a specific compound. In tion 1, and to leave the hydroxyl group at position 2 ready
particular, the carba-L-rhamnose was obtained together to be glycosylated for the synthesis of the above-men-
with a carba-D-gulo compound, starting from a galacto de- tioned trisaccharidic repeating unit.
rivative. Moreover, for the synthesis of the trisaccharidic
repeating unit, we needed a proper pattern of protecting
groups.
As (–)-shikimic acid is quite expensive, it was replaced by
(–)-quinic acid (1) as starting material, which can be eas-
ily converted to (–)-shikimic acid, also allowing the intro-
Looking at the structure and, particularly, at the stereo- duction of desired protecting groups. In a preliminary
chemistry of carba-L-rhamnose, it can be envisaged that attempt, following a literature procedure,7 (–)-quinic acid
such compound shares many features with (–)-shikimic was converted to methyl ester 2, acetylated, dehydrated to
acid (Figure 2, a).
give compound 4, and, after further protecting group ma-
nipulation, converted into 3,4-isopropylidene derivative.
However, the product obtained after reduction of the car-
boxylic ester 6 proved to be too water-soluble making
necessary the introduction of a more hydrophobic protect-
ing group.
reduction
HO
hydroboration-
CH3
OH
oxidation
COOH
CH3
OH
BnO
BnO
ONap
Moreover, although the authors do not comment on the
direction of the elimination reaction, in our hands, the
expected (–)-shikimic ester 4 was accompanied by a sub-
stantial amount of the 4-epi derivative [ratio (–)-shikimic:
(–)-4-epi-shikimic 85:15], which was not separable by
chromatography.
HO
OH
HO
OH
OH
same stereochemistry
a
b
Figure 2 (a) Structural analogies between (–)-shikimic acid and
carba-L-rhamnose; (b) protected carba-L-rhamnose for incorporation
in S. p. 19F repeating unit.
We then decided to prepare the cyclohexylidene deriva-
tive 6, by a modified literature procedure8 in dioxane at
60 °C instead of toluene at reflux. A 0.1 M solution of
compound 5 in dioxane was treated with 3 molar equiva-
lents of cyclohexanone and a catalytic amount (about
10%) of p-toluensulfonic acid for 3 hours. It was now pos-
sible to eliminate the undesired isomer by careful flash
chromatography (Scheme 1).
In particular, the three hydroxyl groups have the same ste-
reochemistry as those in carba-L-rhamnose. Moreover, the
carboxylic acid could be reduced to a methyl group and
the hydroboration–oxidation of the double bond should
allow the installation of the missing hydroxyl group and
the required stereochemistry at the two newly formed
stereocenters. Finally, an appropriate pattern of protecting
groups should be envisaged, in order to synthesize the
Streptococcus pneumoniae 19F repeating unit (Figure 2,
b).
The carboxylic ester 6 was reduced to the corresponding
alcohol 7 with DIBALH9 and the primary allylic hydroxyl
group was selectively replaced for a bromide using tri-
phenylphosphine and carbon tetrabromide10 giving com-
pound 8 in excellent yield (Scheme 2).
Removal of bromine with lithium triethylborohydride11
and protection of the free hydroxyl group as naphthyl-
methyl ether (Nap) in phase-transfer conditions using 18-
crown-6 as catalyst,12 afforded smoothly compound 10,13
ready for the hydroboration–oxidation reaction.
COOMe
COOH
COOMe
HO
HO
HO
ii
i
OAc
OH
AcO
OH
HO
HO
OAc
3
Hydroboration reaction was initially attempted using
BH3·SMe2 complex,14 followed by conventional oxida-
tion with hydrogen peroxide and sodium hydroxide. Al-
though the reaction worked well, the ratio between the
diastereomeric products 11a and 11b was 87:13 (from 1H
NMR).15 We then decided to turn to 9-BBN, which should
be more stereoselective.16 In fact, although it was neces-
sary to work at the reflux of THF, the reaction gave the
desired product 11 in high yield and excellent stereoselec-
tion (11a/11b >97:3, determined by 1H NMR). Protection
of the newly formed hydroxyl group as benzyl ether in the
same conditions previously used for the introduction of
the Nap ether and conventional deprotection of the cyclo-
hexylidene acetal with trifluoroacetic acid gave the diol
13. To introduce regioselectively the benzyl protecting
OH
2
OH
1
iii
COOMe
COOMe
COOMe
v
iv
OH
O
OH
OAc
HO
AcO
O
OH
5
OAc
4
6
Scheme 1 i) MeOH, H2SO4, r.t.; ii) AcCl, pyridine, CH2Cl2, –20 °C;
iii) SOCl2, 0 °C to r.t., 76% over three steps; iv) MeOH, H2SO4, r.t.;
v) cyclohexanone, dioxane PTS, 60 °C, 70% over two steps.
Synlett 2004, No. 14, 2529–2532 © Thieme Stuttgart · New York

LETTER
Transformation of (–)-Quinic Acid into Carba-L-rhamnose
2531
OMe
OH
Br
O
CH3
5
2
1
6
5
4
3
5a
1
2
3
iii
ii
i
4
OH
OH
OH
OH
O
O
O
O
O
CH3
O
O
O
O
9
6
8
7
iv
CH3
CH3
ONap
HO
O
HO
O
5
4
3
5a
1
H3C
HO
v
2
+
O
ONap
ONap
ONap
O
O
O
O
10
11a
11b
vi
O
ONap
ONap
ONap
ONap
viii
H3C
BnO
vii
H3C
BnO
H3C
BnO
H3C
BnO
+
HO
HO
BnO
OBn
14b
OH
OH
O
14a
13
12
Scheme 2 i) DIBALH, THF, 0 °C, 78%; ii) Ph3P, CBr4, syn-collidine, r.t., 93%; iii) SuperHydride®, THF, 0 °C to r.t., 92%; iv) NapBr, KOH,
18-crown-6, THF, r.t., 85%; v) 9-BBN, reflux, then NaOH, H2O2, 0 °C to r.t.; 90% (11a/11b 97:3); vi) BnBr, KOH, 18-crown-6, THF, r.t.,
89%; vii) 60% aq TFA, CH2Cl2, r.t., 87%; viii) BnBr, Bu2SnO, Bu4NBr, reflux, then 50 °C, 96% (14a/14b 93:7).
group on the 3-OH we exploited the stannylene technique,
which should allow the selective introduction of the alky-
lating agent on the equatorial OH group of a 1,2-cis-diol.17
As expected, the reaction gave a 93:7 mixture of the prop-
erly protected carba-L-rhamnose 14a ready for further
glycosylation reaction, together with the 3-OH free com-
pound 14b. The benzylation position of 14a and 14b was
determined after acetylation reaction of the products, on
the basis of the downfield shift of the proton in position 2
or 3, respectively.
References
(1) (a) Fedson, D. S.; Anthony, J.; Scott, G. Vaccine 1999, 17,
S11. (b) Poland, G. A. Vaccine 1999, 17, 1674.
(2) (a) Jones, C. Carbohydr. Eur. 1998, 21, 10. (b) Musher, D.
M. Clin. Infect. Dis. 1991, 14, 801.
(3) See for example: Austrian, R. Vaccine 2001, 19, S71.
(4) See for example: Arndt, B.; Porro, M. In Immunobiology of
Proteins and Peptides VI; Atassi, M. Z., Ed.; Plenum Press:
New York, 1991, 129.
(5) Redlich, H.; Sudau, W.; Szadenings, A. K.; Vollerthun, R.
Carbohydr. Res. 1992, 226, 57.
(6) (a) Gaunt, M. J.; Yu, J.; Spencer, J. B. J. Org. Chem. 1998,
63, 4172. (b) Wright, J. A.; Yu, J.; Spencer, J. B.
Tetrahedron Lett. 2001, 42, 4033. (c) Xia, J.; Abbas, S. A.;
Locke, R. D.; Piskorz, C. F.; Alderfer, J. L.; Matta, K. L.
Tetrahedron Lett. 2000, 41, 169. (d) Liao, W.; Locke, R. D.;
Matta, K. L. Chem. Commun. 2000, 369. (e) Xia, J.;
Alderfer, J. L.; Piskorz, C. F.; Matta, K. L. Chem.–Eur. J.
2001, 7, 356.
(7) Bianco, A.; Brufani, M.; Manna, F.; Melchioni, C.
Carbohydr. Res. 2001, 332, 23.
(8) Fernández, S.; Díaz, M.; Ferrero, M.; Gotor, V. Tetrahedron
Lett. 1997, 38, 5225.
A small amount of compound 11a was also deprotected to
obtain an analytical sample of free carba-L-rhamnose
whose optical rotation was consistent with that described
in the literature.5
Carba-L-rhamnose either unprotected or with a proper
protecting group pattern was obtained in a concise and
high yield way starting from quinic acid. Work is in
progress to synthesize the Streptococcus pneumoniae 19F
repeating unit containing the carbocyclic analog of the
rhamnopyranosidic fragment.
(9) Wang, Z.-X.; Miller, S. M.; Anderson, O. P.; Shi, Y. J. Org.
Chem. 2001, 66, 521.
(10) Lautens, M.; Stammers, T. A. Synthesis 2002, 1993.
(11) Marshall, J. A.; Bourbeau, M. P. Org. Lett. 2002, 4, 3931.
(12) Bessodes, M.; Shamsazar, J.; Antonakis, K. Synthesis 1988,
560.
Acknowledgment
We thank MIUR (COFIN 2002) and Università del Piemonte
Orientale for financial support.
(13) All new compounds have been characterized by proton and
carbon NMR, exact mass, and/or elemental analysis.
Spectroscopic data for selected compounds:
5a-Carba-2,3-O-cyclohexylidene-1-O-naphtylmetyl-a-L-
rhamno-pyranoside (11a): 1H NMR (300 MHz, CDCl3):
d = 1.05 (d, 3 H, JMe,5 = 6.0 Hz, Me), 1.30–1.80 (m, 12 H,
Synlett 2004, No. 14, 2529–2532 © Thieme Stuttgart · New York

2532
E. Valsecchi et al.
LETTER
H-5a and 5 CH2), 1.80–2.00 (m, 1 H, H-5), 3.30 (dd, 1 H,
C-5), 32.05 (t, C-5a), 69.74 (d), 71.20 (t), 72.69 (t), 75.43 (t),
76.16 (d), 82.33 (d), 83.32 (d), 125.71 (d), 125.96 (d), 126.15
(d), 126.25 (d), 127.64 (d), 127.83 (d), 127.92 (d), 127.99
(d), 128.02 (d, 2 C), 128.13 (d, 2 C), 128.24 (d), 128.47 (d, 2
C), 128.60 (d, 2 C), 133.08 (s), 133.40 (s), 136.20 (s), 138.51
(s), 139.14 (s). [a]D25 –18.3 (c = 1, CHCl3).
J3,4 = 7.7 Hz, J4,5 = 9.9 Hz, H-4), 3.90 (br m, 1 H, H-1), 4.05
(dd, 1 H, J2,3 = 5.8 Hz, J3,4 = 7.7 Hz, H-3), 4.29 (br m, 1 H,
H-2), 4.72 and 4.78 (ABq, 2 H, CH2-Nap), 7.20–7.90 (m, 7
H, Ar H). 13C NMR (75.4 MHz, CDCl3): d = 18.21 (q, CH3),
23.81 (t), 24.17 (t), 25.13 (t), 30.06 (d, C-5), 32.60 (t), 35.41
(t), 38.10 (t), 71.32 (t), 74.15 (d), 76.71 (d), 78.12 (d), 80.22
(d), 109.74 (s), 125.99 (d), 126.24 (d, 2 C), 127.79 (d),
127.96 (d), 128.31 (d, 2 C), 133.09 (s), 133.37 (s), 135.77
(s). [a]D25 +15.7 (c = 1, CHCl3). Mp 99.0–101.0 °C.
5a-Carba-1-naphtylmetyl-3,4-dibenzyl-a-L-rhamno-
pyranoside (14a,b): 1H NMR (300 MHz, CDCl3): 1.10 (d, 3
H, J6,5 = 6.6 Hz, H-6), 1.62 (dt, 1 H, J5aA,5aB = J5aA,5 = 14.5
Hz, J5aA,1 = 2.5 Hz, H-5aA), 1.86 (dt, 1 H, J5aA,5aB = 14.5 Hz,
(14) Takahashi, T.; Kotsubo, H.; Iyobe, A.; Namiki, T.; Koizumi,
T. J. Chem. Soc., Perkin Trans. 1 1990, 3065.
(15) The stereochemistry of compounds 11a and 11b was
inferred from the coupling constants of the signal of H-C(4)
in 1H NMR spectrum. In fact, H-C(4) appeared as a doublet
of doublets for both compounds, but in 11a the 3J were 7.7
Hz and 9.9 Hz, while in compound 11b the constants were
10.8 Hz and 3.2 Hz. Furthermore, H-C(4) in compound 13
appeared as a triplet with a 3J of 9.9 Hz, clearly indicating the
axial orientation of H-C(4) in a chair conformation of 13.
(16) Shing, T. K. M.; Cui, Y.; Tang, Y. Tetrahedron 1992, 48,
2349.
J5aB,5 = J5aB,1 = 2.7 Hz, H-5aB), 1.97–2.17 (m, 1 H, H-5),
2.58 (br s, 1 H, -OH), 3.39 (t, 1 H, J3,4 = J4,5 = 9.1 Hz, H-4),
3.80 (br m, 1 H, H-1), 3.86 (dd, 1 H, J2,3 = 3.3 Hz, J3,4 = 9.1
Hz, H-3), 4.19 (br t, 1 H, H-2), 4.55–5.00 (m, 6 H, 3 ABq,
CH2-Nap and CH2-Bn), 7.10–8.00 (m, 17 H, Ar H). 13
C
(17) David, S.; Hanessian, S. Tetrahedron 1985, 41, 643.
NMR (75.4 MHz, CDCl3): d = 18.34 (q, C-6), 31.78 (d,
Synlett 2004, No. 14, 2529–2532 © Thieme Stuttgart · New York